A statistical approach for water movement in the unsaturated zone
Zhang, Tielin LU
(1991)
In Report
1010.
- Abstract
- This work presents a statistical approach for estimating and analyzing the downward transport pattern and distribution of soil water by the use of pattern analysis of space-time correlation structures. This approach, called the Space-time Correlation Field, is mainly based on the analyses of correlation functions simultaneously in the space and time domain.
The overall purpose of this work is to derive an alternative procedure in soil moisture analysis without involving detailed information on hydraulic parameters and to visualize the dynamics of soil water variability in the space and time domains. The basics of this work consist of the development, verification, and application of two models; one numerical and the other statistical.... (More) - This work presents a statistical approach for estimating and analyzing the downward transport pattern and distribution of soil water by the use of pattern analysis of space-time correlation structures. This approach, called the Space-time Correlation Field, is mainly based on the analyses of correlation functions simultaneously in the space and time domain.
The overall purpose of this work is to derive an alternative procedure in soil moisture analysis without involving detailed information on hydraulic parameters and to visualize the dynamics of soil water variability in the space and time domains. The basics of this work consist of the development, verification, and application of two models; one numerical and the other statistical. The major purpose of the numerical model is to calibrate and verify the statistical model, which is then applied to field measured soil water data to analyze the patterns and scales of soil water variability.
The numerical model employs the method of characteristics to solve the one-dimensional unsaturated flow equations for different boundary and initial conditions. Different treatments and combinations of the soil hydrological parame¬ters are used in the numerical simulations to generate various patterns of hypothetical time series. The output from the numerical modelling is used as input in the application of the space-time correlation field technique, which is, after the verification and calibration, applied to the field measured data. The field measure¬ment of soil water involves two different catchments in Southern Sweden and the measured time series range from daily to monthly during periods lasting up to ten year. Basic statistical analyses of soil water variability are carried out for the field-measured data. These include temporal and spatial variability analyses using the coefficient of variation, cross-correlation and semi-variograms.
The results of the application show that the space-time correlation fields display effects of soil layers with different hydraulic properties and boundaries between them. Besides soil parameters and soil structure, the rainfall pattern strongly affects the patterns of correlation isolines. Field soil heterogeneities are displayed in the space-time correlation fields simultaneously. Even very close sites are shown to have quite different space-time correlation fields, indicating a small-scale spatial variability of hydraulic properties. It is concluded that the approach poses special advantages when visualizing time and space dependent properties simultaneously.
The most important results and conclusions of this work are:
∙The statistical approach presented in this study can be used to investi¬gate the hydrological response of soil water dynamics and characteris¬tics in different dimensions (space and time) and scales.
∙Soil water content and soil physical properties are temporally and spatially dependent. For the two catchment areas studied, it is shown that the vertical variability is more significant than the horizontal in the correlation structure, although micro-scale horizontal variability does exist. A larger scale of horizontal dependence seems to be present, as indicated by the semi-variogram analysis even though the correlograms show a random scatter. Temporal variability is mostly greatest in the upper layers of the soil profile.
∙The patterns of the space-time correlation fields reflect the combinations of soil physical parameters (unsaturated hydraulic conductivity and dispersion coefficient). It is shown that the different combinations of these parameters alter the orientations and patterns of the correspond¬ing correlation fields. This can be used to identify the dominant component in unsaturat¬ed flow systems.
∙The space-time correlation fields can visualize effects of changes in infiltrated input (e.g., rainfall). It is possible to estimate the pattern and the propagation rate downwards of moisture movement in the soil profile, especially for the case when a pulse of rain water infiltrates down towards the groundwater table.
∙Small-scale soil heterogeneities vertically and horizontally can be identified by the correlation field. This is especially true when the layered soil has sharp changes in its physical and/or hydraulic proper¬ties such as hydraulic conductivi¬ty and dispersion coefficient.
Since the correlation field technique gives a statistical measure of the dependent property that varies within the space-time field, it is possible to interpolate the fields to points where observations are not available, estimating spatial or temporal averages from discrete observations. The advantage of using the correlation field method instead of directly measured time series in variability analysis is that the heterogeneity and space-time dependence are often hidden in the observed original time series.
Keywords:Soil moisture, Space-time correlation, Numerical modelling, Scale effect, Variability (Less)
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/159075b2-a0f1-4443-8e3d-4b5157789db2
- author
- Zhang, Tielin
LU
- supervisor
-
- Gunnar Lindh LU
- organization
- publishing date
- 1991
- type
- Thesis
- publication status
- published
- subject
- in
- Report
- volume
- 1010
- pages
- 250 pages
- publisher
- Department of Water Resources Engineering, Lund University
- ISSN
- 1101-9824
- language
- English
- LU publication?
- yes
- id
- 159075b2-a0f1-4443-8e3d-4b5157789db2
- date added to LUP
- 2025-02-16 12:06:54
- date last changed
- 2025-04-08 16:10:28
@phdthesis{159075b2-a0f1-4443-8e3d-4b5157789db2,
abstract = {{This work presents a statistical approach for estimating and analyzing the downward transport pattern and distribution of soil water by the use of pattern analysis of space-time correlation structures. This approach, called the Space-time Correlation Field, is mainly based on the analyses of correlation functions simultaneously in the space and time domain.<br/>The overall purpose of this work is to derive an alternative procedure in soil moisture analysis without involving detailed information on hydraulic parameters and to visualize the dynamics of soil water variability in the space and time domains. The basics of this work consist of the development, verification, and application of two models; one numerical and the other statistical. The major purpose of the numerical model is to calibrate and verify the statistical model, which is then applied to field measured soil water data to analyze the patterns and scales of soil water variability.<br/>The numerical model employs the method of characteristics to solve the one-dimensional unsaturated flow equations for different boundary and initial conditions. Different treatments and combinations of the soil hydrological parame¬ters are used in the numerical simulations to generate various patterns of hypothetical time series. The output from the numerical modelling is used as input in the application of the space-time correlation field technique, which is, after the verification and calibration, applied to the field measured data. The field measure¬ment of soil water involves two different catchments in Southern Sweden and the measured time series range from daily to monthly during periods lasting up to ten year. Basic statistical analyses of soil water variability are carried out for the field-measured data. These include temporal and spatial variability analyses using the coefficient of variation, cross-correlation and semi-variograms.<br/>The results of the application show that the space-time correlation fields display effects of soil layers with different hydraulic properties and boundaries between them. Besides soil parameters and soil structure, the rainfall pattern strongly affects the patterns of correlation isolines. Field soil heterogeneities are displayed in the space-time correlation fields simultaneously. Even very close sites are shown to have quite different space-time correlation fields, indicating a small-scale spatial variability of hydraulic properties. It is concluded that the approach poses special advantages when visualizing time and space dependent properties simultaneously.<br/>The most important results and conclusions of this work are:<br/><br/>∙The statistical approach presented in this study can be used to investi¬gate the hydrological response of soil water dynamics and characteris¬tics in different dimensions (space and time) and scales.<br/><br/>∙Soil water content and soil physical properties are temporally and spatially dependent. For the two catchment areas studied, it is shown that the vertical variability is more significant than the horizontal in the correlation structure, although micro-scale horizontal variability does exist. A larger scale of horizontal dependence seems to be present, as indicated by the semi-variogram analysis even though the correlograms show a random scatter. Temporal variability is mostly greatest in the upper layers of the soil profile.<br/><br/>∙The patterns of the space-time correlation fields reflect the combinations of soil physical parameters (unsaturated hydraulic conductivity and dispersion coefficient). It is shown that the different combinations of these parameters alter the orientations and patterns of the correspond¬ing correlation fields. This can be used to identify the dominant component in unsaturat¬ed flow systems.<br/><br/>∙The space-time correlation fields can visualize effects of changes in infiltrated input (e.g., rainfall). It is possible to estimate the pattern and the propagation rate downwards of moisture movement in the soil profile, especially for the case when a pulse of rain water infiltrates down towards the groundwater table.<br/><br/>∙Small-scale soil heterogeneities vertically and horizontally can be identified by the correlation field. This is especially true when the layered soil has sharp changes in its physical and/or hydraulic proper¬ties such as hydraulic conductivi¬ty and dispersion coefficient.<br/><br/>Since the correlation field technique gives a statistical measure of the dependent property that varies within the space-time field, it is possible to interpolate the fields to points where observations are not available, estimating spatial or temporal averages from discrete observations. The advantage of using the correlation field method instead of directly measured time series in variability analysis is that the heterogeneity and space-time dependence are often hidden in the observed original time series.<br/><br/>Keywords:Soil moisture, Space-time correlation, Numerical modelling, Scale effect, Variability}},
author = {{Zhang, Tielin}},
issn = {{1101-9824}},
language = {{eng}},
publisher = {{Department of Water Resources Engineering, Lund University}},
school = {{Lund University}},
series = {{Report}},
title = {{A statistical approach for water movement in the unsaturated zone}},
volume = {{1010}},
year = {{1991}},
}